1. Field of the Invention
The present invention relates to an image processing apparatus, an image processing method and a computer readable recording medium storing a program.
2. Description of Related Art
Conventionally, in image forming apparatuses which form images on recording media by using toner or ink, halftone images have been formed by screen processing in order to express half tones. As the screen processing, there is widely used a systematic dither method for performing binarization on the basis of large/small relationships between pixel values of a bitmap image and the respective threshold values regulated by threshold value matrices.
The systematic dither method includes AM modulation (halftone dot screen) for changing sizes of halftone dots according to gradation and FM modulation (FM dither method) for changing the number (frequency) of halftone dots to be dotted, each of the methods being used for image formation. However, image forming apparatuses using the methods do not necessarily output color dots of respective pixels isotropically due to the difference in characteristics according to the image formation, especially, due to the difference in characteristics according to the output in a conveyance direction (sub scanning direction) and its vertical direction (main scanning direction) of recording medium. Accordingly, there are generated portions where halftone dots are crushed or interrupted together in specific positions or directions due to the arrangement of threshold values in threshold value matrices. Since a threshold value matrix is normally applied repeatedly in a tiled manner to image data which is a forming target, as a result, there is a problem (artifacts) that artificial patterns are visually recognized in the output image due to the non-uniformity in density being repeatedly generated with an application cycle of dither matrix.
As for the problem, Patent Document 1 (Japanese Patent Application Laid Open Publication No. 2013-17109) discloses a technique for reducing cyclic visual recognition of crush and void of halftone dots by setting arrangement of threshold values in the main scanning direction and the sub scanning direction of a dither matrix with a frequency ratio according to the ratio of line width of thin line in the main scanning direction to line width of thin line in the sub scanning direction which were actually measured, and preventing the output overlapping patterns for each color from being repeated every cycle by making the sizes of dither matrices different from each other for the respective colors of color output.
On the other hand, the patterns which appear in a cyclic way vary in the degree of profile (visibility level of patterns) depending on various conditions. One of the various conditions is spatial period dependency of patterns. Accordingly, there is a conventional method for reducing visibility by appropriately adjusting sizes of dither matrices.
However, such visibility of artifacts can vary depending on combination of the above-mentioned various conditions. Then, conventional configurations have a problem that allocation amounts of memories corresponding to the sizes of threshold matrices are fixed and it is difficult to set threshold value matrices flexibly and easily.